[rabbitmq-discuss] active/active HA and publishers/consumers

[Steve Martin]

It suddenly clicked, thank you.
I had in my mind the model of database log-shipping, which was incorrect.

On a tangential albeit related note, can anyone suggest an approach for
rigging up a load balancer in a development environment in order that I may
test out my application and HA cluster?
I currently have a number of VMs running RMQ in the cluster (several
instances of VMWare player running Ubunto 10.10 as I preferred modelling the
ultimate production scenario rather than setting up a single machine
cluster)
Is there a simple load balancing virtual appliance that can be used in the
same manner, and will this open up a networking can of worms? (I'm
predominately a software tech rather than a networker)

Steve

On 8 September 2011 15:03, Matthew Sackman <matthew at rabbitmq.com> wrote:

> On Thu, Sep 08, 2011 at 06:40:19AM -0700, Steve Martin wrote:
> > I, too, would like to echo John's initial query.
> > I have an active/active HA cluster and would like to seek some
> > guidance on how best to configure my system with regards to how to
> > know where to publish to/consume from.
> > If the master fails and a slave is promoted, how do I know what the
> > new master is or can I publish to the slave (my thoughts are no but I
> > have to ask)?
>
> You can publish from _any_ node in the cluster. It doesn't matter. The
> nodes in the cluster know where the master is and will forward
> publications internally.
>
> Matthew
> _______________________________________________
> rabbitmq-discuss mailing list
> rabbitmq-discuss at lists.rabbitmq.com
> https://lists.rabbitmq.com/cgi-bin/mailman/listinfo/rabbitmq-discuss
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rabbitmq.com/pipermail/rabbitmq-discuss/attachments/20110908/9575b03a/attachment.htm>